Atrial fibrillation (AF) is the most common diagnosed arrhythmia, with the number Americans diagnosed with AF estimated at 2.2 million and continuing to increase as the population ages. Electrical activity during AF is characterized by complex activation patterns that are difficult to map. However, there is growing evidence that AF is driven by stable sources in the atria, such as rotors or focal activity. The most widely used ablation and surgical interventions to treat patients with AF use generalized, empiric strategies that target that do not consider the specific mechanisms of the arrhythmia. Thus, methods utilizing AF electrograms that effectively locate AF sources in the atria are highly sought after. We have recently developed an electrogram morphology recurrence analysis that has been able to identify areas of the atria that have high rates of electrogram morphologies recurrence. Preliminary study has shown that patients who underwent ablation targeting the pulmonary veins, electrograms with highest recurrence rates located in the left atrium predicted ablation success. In contrast, having these fast/recurring electrograms in the right atrium indicated a very poor ablation success rate. The overall aim is to evaluate the pathophysiological significance of electrograms with high recurrence rates. In the first aim, we will test whether sites with highest rates of electrogram morphology recurrence uniquely represent area of AF sources in a canine model of AF by performing ablation on those sites. Pulmonary vein ablation will be performed in a control group as a comparison. In the second aim, we will assess the anatomic and autonomic characteristics of sites with the highest morphology recurrence rates using the canine model of AF. In the third aim, we will compare the effect of pulmonary vein ablation in patients with and without sites of high morphology recurrence rates located near the veins. If high morphology recurrence rates represent driver activity, as is hypothesized, then we expect AF termination or cycle length slowing. In the fourth specific aim, we will assess the ability of surface ECG mapping and MRI to predict the atrium (right or left) that contains the site of highest rate of electrogram morphology recurrence. Completion of all the aims will provide a clear path to using recurrence mapping as a new clinical tool to guide ablation or surgery interventions of AF.